Validación de un método de adquisición de la condición corporal en vacas de cría

The main objective of this work was the validation of a method of acquiring body condition in breeding cows. It was based on developing a methodology that would make it possible the determination of the body condition (CC) of AA cows in an objective way, through qualifications of this form images archived by the Cond – Corp device. The operation of the corresponding components was validated, both, the reader, and the device (connected to each other). Through correlation result, it was established that it is possible to determine the level of CC from the observation of images taken by the device. The experiment was carried out in “La Casona”, Casupá, Florida department. The animals used for the experiments were thirty Aberdeen Angus cows, black and red. This work was divided into two stages: the first in the country and the second one, on desk. At the beginning of the practical work, the first cow was used to understand the operation of the components, adjust the height of the placement of the device and, the external memory was extracted, placing it on a computer to corroborate the file. Then, three cows per tube were worked for a better comfort, we proceeded to cap, since, when analyzing the images obtained, due to the movement of the animals, an adequate sharpness was not archived to be able to classify in the next stage of the work. At the end of the tube, a trained observer was placed to rate the body condition of each cow, making it possible to obtain a form with the assessment of each one. The second stage consisted of qualifying the body condition at desk with the images obtained by the Cond- Corp device. To do this, the external memory was extracted, placed on the computer, and the ten photos taken by the device of each cow were downloaded. Once the results were obtained, the correlation of both evaluations collected were calculated, the variation within each cow in the ten photos was analyzed. The correlation coefficient between both methods of estimation was 0,91. Also, the coefficient of determination (r2) was calculated, which was 0,83, while the CC cond-corp model= 0,33 + 0,88 x, indicating that, for each unit of change in the CC of the expert (visual appreciation) CC to desktop changes by 0,88. Finally, the exact coincidences and the differences found of ±0,25 between the ratings of the device and the photos taken by the cellphone were evaluated according to the expert (reference). The results obtained from the device were better, superior exact matches were obtained and, at the same time, the total number of cows that were analyzed none one overstep 0,25, as was the case with the classification from the photos taken by the cell phone.